Molecular profiling of renal medullary carcinoma to reveal frequent alterations in chromatin remodeling genes and to identify EZH2 as a relevant therapeutic target.

Abstract

571 Background: Renal medullary carcinoma (RMC) is a rare and highly aggressive malignancy that affects primarily young adults with sickle cell trait. The prognosis of RMC is poor with a median survival of about 12 months from diagnosis. It has been previously shown that loss of heterozygosity of SMARCB1 occurs in the majority of cases leading to protein loss. However, a comprehensive molecular profiling of this disease remains unknown. Methods: Whole-exome sequencing (WES) of 9 RMC matched tumor-normal cases and 1 unmatched case were performed. In addition, single nucleotide polymorphism (SNP) analysis was done using the Affymetrix SNP 6.0 array. To test differentially expressed genes between RMC cases and normal kidney, RNA sequencing was carried out. The effect of EZH2 inhibition on in vitro RMC tumor cell proliferation was assessed using the MTT assay. Results: WES analysis of 9 paired RMC cases identified 233 non-synonymous somatic mutations (median: 21; range: 12-44). We observed frequent mutations in chromatin remodeling genes in 7 of 10 RMC cases, including mutations in SMARCB1 (n = 2) and SMARCA5 gene (n = 1). SNP analysis demonstrated that only the case with SMARCA5 somatic mutation harbored SMARCB1 monoallelic loss. However, all RMC cases had loss of SMARCB1 protein by immunohistochemistry. Ingenuity pathway analysis using RNA sequencing data identified EZH2 as one of the top over-expressed upstream regulators (Z-score = 3.7; p-value = 1.6x10-6). RT-PCR and western blot analyses confirmed that EZH2 expression is elevated at both mRNA and protein levels corresponding to SMARCB1 loss in RMC tumors. In vitro assessment of a novel RMC tumor cell line derived from the case harboring SMARCB1 deletion showed that EZH2 inhibitor GSK343 inhibited proliferation of this cell line. Conclusions: Our data reveal that 30% of RMC cases showed inactivation of SMARCB1 through somatic mutations or monoallelic loss, suggesting that additional epigenetic mechanisms may lead to gene inactivation. In addition, our data suggest that EZH2 may be a relevant therapeutic target in RMC.